The Quiet Number That Moves the Money.

A earlier this year a trade publication ran a piece built around something I had written on LinkedIn “LHV vs HHV in Data Centers: Why Fuel Definitions Matter for Onsite Power Economics”. The subject was not a new gas engine, a funding round, or a gigawatt-scale announcement. It was the difference between two ways of measuring the energy in natural gas. Lower heating value and higher heating value. The kind of distinction that usually lives in an engineering appendix, three layers below anything anyone calls a story.

I have been thinking about why it got picked up at all.

The honest answer is that the definitions themselves have not changed. Gas engines have always been rated on an LHV basis, because they cannot recover the latent heat stored in water vapor. Gas in the United States and most other countries has always been billed on an HHV basis, because that is the total energy content the supplier delivers. The ratio between the two has sat at roughly 1.108 for as long as anyone has been running these machines. None of that is news. What changed is the context the numbers now sit inside.

For most of the last twenty years, this gap was a reconciliation problem for specialists. A site engineer knew to size the engine on LHV and forecast the fuel bill on HHV, and the rest of the business never had to look at it. The two perspectives were quietly aligned by people who did this for a living, and the alignment never surfaced as a decision because the stakes were small enough to absorb.

That is no longer true, and the reason is the data center.

When a campus moves from drawing power off the grid to generating it onsite at scale, the fuel definition stops being a footnote and becomes a line item. A few points of apparent efficiency, multiplied across multi-megawatt loads running thousands of hours a year, compounds into a number that changes how a project pencils out. Worse, the error is usually invisible. When one technology is evaluated on HHV and another on LHV, often without anyone noticing the mismatch, the comparison looks clean and is quietly wrong. Gas engines can be made to look less efficient than they are. Operating expense models can be built on the wrong basis. PUE and sustainability reporting can drift away from reality. The mistake does not announce itself. It just sits in the spreadsheet, moving capital toward the wrong answer.

This is what I find worth reflecting on. The piece did not get attention because the engineering was novel. It got attention because the audience for the engineering has changed. People who used to delegate this entirely are now the ones signing off on onsite generation, and they are discovering that the unglamorous conventions are the ones that move money. The market is catching up to something the specialists always knew, and the catching up is happening because the cost of getting it wrong has crossed a threshold.

I think this is a pattern, not a one-time event. As more compute capacity is built behind the meter, a whole category of formerly invisible engineering knowledge is going to surface into commercial visibility. Gas flow sizing and the headroom you need for phased expansion. Pressure regulation and metering. The redundancy math that decides whether a contracted gas limit becomes a constraint two years after commissioning. These are not exotic topics. They are the load-bearing details that have always governed whether onsite power actually delivers what the model promised. They are simply being read now by people who were not reading them before.

The strategic read, for anyone building in this space, is that the next phase of competition is not only about who can deliver power fastest. It is about who can model it honestly. The teams that align their assumptions, that know which basis governs sizing and which governs cost, that can explain why their efficiency number means what it says, will produce projections that hold up over a fifteen-year life. The teams that cannot will keep generating confident numbers that quietly disagree with the physical plant.

That a fuel definition became news, even briefly, tells me the market is starting to ask the right questions. The work now is to make sure the answers are built on the right basis before the capital commits to them.

Questions and answers

What is the difference between LHV and HHV?

Higher heating value (HHV) is the total energy content of a fuel, including the latent heat held in the water vapor produced when it burns. Lower heating value (LHV) excludes that latent heat and represents the energy actually available to do work in an engine. For natural gas the two are related by a ratio of roughly 1.108, with HHV being the larger figure.

Why are gas engines rated on an LHV basis?

Because a gas engine (or turbine) cannot recover the latent heat in the exhaust water vapor, that energy never becomes mechanical or electrical output. Stating efficiency on an LHV basis reflects the energy the engine can actually convert. This holds whether the engine runs as prime power, in grid-parallel operation, or as standby within a microgrid.

If engines run on LHV, why does HHV matter for a data center?

Because that is how the fuel is billed. In the United States, United Kingdom and more, gas suppliers price natural gas on an HHV basis, and that is the number that appears in supply contracts, utility bills, and operating expense models. So a project lives in two measurement systems at once: sizing and efficiency are governed by LHV, while fuel cost is governed by HHV. Reliable long-term projections depend on using each one where it belongs.

What goes wrong when the two are mixed?

The comparison looks clean and is quietly wrong. If one technology is evaluated on HHV and another on LHV, often without anyone noticing, gas engines can appear less efficient than they are, operating expense forecasts can be built on the wrong basis, and PUE or sustainability reporting can drift from reality. The error does not announce itself. It sits in the model and moves capital toward the wrong answer.

Does the fuel definition affect more than cost?

Yes. LHV also determines the volumetric gas flow rate needed to support a given electrical load. If the heating value of the supplied gas changes, the flow rate has to change with it. That feeds directly into gas supply capacity and redundancy, pressure regulation and metering, and the contracted gas limits that decide how much expansion headroom a site actually has. For campuses planning phased or modular growth, it is an infrastructure question, not just an accounting one.

Why is this becoming a more visible issue now?

The definitions have not changed. The audience has. As data centers move from drawing grid power to generating it onsite at scale, decisions that specialists used to handle quietly are now being signed off by people closer to the capital. The unglamorous conventions turn out to be the ones that move money, and the market is catching up to what engineers have always known because the cost of getting it wrong has crossed a threshold.